Create LSTM.py

LSTM.py

@@ -0,0 +1,46 @@
+# Importing required libraries
+import numpy as np
+import tensorflow as tf
+from tensorflow.keras.models import Sequential
+from tensorflow.keras.layers import LSTM, Dense
+from sklearn.preprocessing import MinMaxScaler
+
+# Generate some example data (sine wave sequence prediction)
+def generate_sine_data(seq_length, num_samples):
+    X, y = [], []
+    for _ in range(num_samples):
+        start = np.random.rand()
+        x_seq = np.sin(np.linspace(start, start + 2 * np.pi, seq_length))
+        X.append(x_seq)
+        y.append(np.sin(start + 2 * np.pi))  # Next value in sequence
+    return np.array(X), np.array(y)
+
+# Hyperparameters
+sequence_length = 50
+num_samples = 1000
+batch_size = 32
+epochs = 50
+
+# Generate and scale the data
+X, y = generate_sine_data(sequence_length, num_samples)
+scaler = MinMaxScaler(feature_range=(0, 1))
+X_scaled = scaler.fit_transform(X)
+X_scaled = np.reshape(X_scaled, (X_scaled.shape[0], X_scaled.shape[1], 1))  # Reshape for LSTM
+
+# Building the LSTM model
+model = Sequential()
+model.add(LSTM(50, activation='relu', input_shape=(sequence_length, 1)))
+model.add(Dense(1))
+
+# Compile the model
+model.compile(optimizer='adam', loss='mse')
+
+# Train the model
+model.fit(X_scaled, y, epochs=epochs, batch_size=batch_size)
+
+# Example prediction
+test_input = np.sin(np.linspace(0, 2 * np.pi, sequence_length))
+test_input_scaled = scaler.transform(test_input.reshape(1, -1))
+test_input_scaled = np.reshape(test_input_scaled, (test_input_scaled.shape[0], test_input_scaled.shape[1], 1))
+predicted_value = model.predict(test_input_scaled)
+print(f"Predicted next value: {predicted_value[0][0]}")